College entrance examinations and race relations

American colleges and universities consider a variety of factors when selecting students for admission. These factors may include, but are not necessarily limited to, high school grades, class rank, difficulty of courses taken, personal interviews, letters of reference, and samples of students’ written works. In addition, most colleges and universities require students to submit test scores from one or more of several nationally administered standardized tests. Known collectively as college entrance examinations, the most commonly used tests in the United States are the American College Test (ACT) and the Scholastic Aptitude Test (SAT). Some students also may take Achievement Tests (ATs) and the Preliminary Scholastic Aptitude Test/National Merit Scholarship Qualifying Test (PSAT/NMSQT).

The Tests

According to the American College Testing Assessment Program (ACTAP), which created and administers the ACT, the test does not ask students to recall specific information learned in high school but rather asks students to demonstrate their reasoning ability in four fields: English, mathematics, reading, and scientific reasoning. Students receive scores on a scale of 1 (low) to 36 (high) for each of the four skill areas, and a composite score. In addition, students receive scores on a scale of 1 (low) to 18 (high) for each of the seven subsections of the test. Student scores are also presented as percentiles.

The SAT, created and developed by Educational Testing Service (ETS), is administered in cooperation with the College Entrance Examination Board. Like the ACT, the SAT is designed to measure skills necessary for college-level work. The SAT tests two basic skill areas, verbal ability and mathematical reasoning, using a multiple-choice format.

The version of the SAT in use until the mid-1990s had two verbal sections containing questions involving antonyms, analogies, sentence completion, and reading comprehension. It also had two mathematical reasoning sections containing questions involving problem solving and quantitative comparisons. Each of these skill areas (verbal ability and mathematical reasoning) was scored on a scale of 200 (low) to 800 (high). Composite and percentile scores were provided. This version of the SAT also contained a fifty-question test of standard written English and a series of experimental questions. The experimental questions are not included in the student’s score but are used in the development of future test questions.

ATs, administered by the College Board, were examinations based on specific knowledge of subjects including American history, European history, French, German, Hebrew, Latin, Spanish, mathematics, biology, chemistry, physics, English composition, and English literature. Some colleges required students to take these tests in addition to the SAT.

In the 1990s, Educational Testing Service developed new tests, the SAT I and SAT II, to replace the SAT. Like the SAT, the SAT I tests two basic skill areas: verbal and mathematical reasoning and is scored similarly. The SAT I verbal reasoning section focuses on analogies, sentence completion, and critical reasoning and contains a new critical reading section that tests vocabulary in context, analysis and synthesis, interpretation, and evaluation. The mathematical reasoning section allows the use of calculators, and a proportion of examination questions require students to enter the answer they calculate. In 2025, the test moved to a shorter digital version which included an embedded graphing calculator.

The SAT II was designed to replace the ATs and the test of standard written English. Content areas are expanded, and the test of standard written English is combined with the English composition test (an AT). The new writing test of the SAT II contains a series of multiple-choice questions testing grammar and usage and an essay section. In 2021, the SAT II was discontinued.

The Preliminary Scholastic Aptitude Test/National Merit Scholarship Qualifying Test, administered primarily during the junior year in high school, contains two sections testing verbal skills and mathematical reasoning in a multiple-choice format. Test scores are used to award National Merit Scholarships, National Achievement Scholarships, and Achievement Scholarships.

Criticisms of the Tests

Despite their importance and wide use, college entrance examinations are not without their flaws or their critics. The two main organizations that create college entrance examinations, Educational Testing Service and the American College Testing Assessment Program, work to maintain and improve the validity and reliability of the tests and to avoid bias. Nevertheless, the tests have been subjected to some serious criticisms, and it is because of the tests’ importance that such criticism must be considered. Criticisms focus on the existence of bias and flawed questions, testing procedures, and accusations of lack of due process, violation of privacy, and misuse of test results. Critics have argued that standardized college entrance examinations are biased against women, the poor, minorities, students from rural areas, and students whose first or primary language is not English.

Other things being equal, high test scorers should perform better in college than low test scorers. National SAT test score averages have consistently showed that men perform better on the test than women. The gender gap in test scores varies by year but is generally 50 to 60 points. Gender differences on the SAT are greater on the mathematical reasoning section than on the verbal ability section. A similar pattern exists regarding PSAT/NMSQT scores. Men also score higher on the mathematics, reading, and scientific reasoning sections of the ACT. If standardized college entrance examinations accurately predict future academic success, men should have higher average grades in college than women. Yet, at least during the first year of college, women have higher average grades than men.

Educational Testing Service states that women’s scores are lower because greater numbers of women have been taking the test since 1970 than took it in previous decades. Consequently, the test pool for women is increasingly less selective and includes a greater number of low scorers, reducing the overall average. Ruth Ekstrom, Marilane Lockheed, and Thomas Donlon, however, have found that the structure of test questions influences gender performance. Women are less likely to answer a test item correctly if the question contains only, or refers primarily to, male characters. A survey of the reading comprehension passages contained in SAT examinations showed that 93 percent of the characters to which the passages refer are male. In addition, evidence indicates that college entrance examinations place more emphasis on subject areas such as science and mathematics in which men have traditionally outperformed women. Accordingly, test content selection may account for some of the gender differences in test scores.

Race/EthnicityTotalMenWomenAmerican Indian 3.1 3.0 3.2Black10.6 9.911.1White80.780.780.7Mexican American 1.5 1.4 1.6Puerto Rican 1.1 1.2 1.0Other Latino 1.6 1.7 1.6Chinese American 1.4 1.6 1.2Filipino American 0.8 0.8 0.8Japanese American 0.4 0.4 0.4Korean American 0.7 0.7 0.7Southeast Asian American 0.6 0.7 0.6Other Asian American 1.4 1.5 1.3Other 2.9 2.9 2.8

Students from wealthier families score higher on college entrance examinations. In the case of the SAT, students whose family income is more than $70,000 a year have test-score averages of 996. Students whose family income is under $10,000 a year have test-score averages of 780. ACT test scores and family income show a similar correlation. These patterns may stem from differences in educational opportunities rather than differences in aptitude. Additionally, less affluent students may be unable to afford test preparation courses.

Minority students score lower on college entrance examinations than White students. Differences vary by group and over time. For example, 1989 SAT scores show that average combined test scores for African Americans were 200 points lower than those of White students. Of all minority groups, Asian Americans performed the best. The 1989 SAT average combined score for Asian Americans was 934, compared to 937 for White students. ACT average composite scores follow similar patterns. Almost forty years later and after changes to the tests based on criticisms, results did not change much. In 2023, SAT scores showed an average of 908 for Black students compared to 1082 for White students and 1219 for Asian students.

There is also evidence that college entrance examinations may contain bias against students from rural areas and students whose first or primary language is not English. James Loewen found that incoming University of Vermont students from rural areas have SAT scores that average 100 points lower than those of students from urban areas; however, the actual academic performance of the two groups is similar. Alicia Schmitt found that Hispanic students who take the SAT are much more likely to answer incorrectly questions that contain cognates and/or homographs. Consequently, differences in test scores between Latinos and whites may be partially the result of language differences, not differences in aptitude for college work.

In addition to bias, critics have charged that college entrance examinations contain flawed questions (questions that have more than one meaning or more than one answer), which may unfairly penalize test takers. Critics have also charged ETS with using unethical testing procedures, lack of due process protection, and invasion of privacy. John Weiss, Barbara Beckwith, and Bob Schaeffer claim the experimental questions on the SAT are fundamentally unfair because they violate the principle of informed consent and are flawed. In addition, no due process protection exists for students who are accused of cheating on the examination. Students may appeal to ETS, but ETS has sole discretion in determining guilt or innocence in cases of suspected cheating.

Finally, college entrance examinations have been criticized for being misused by scholarship agencies and colleges and universities. Many scholarships are awarded primarily on the basis of PSAT/NMSQT, SAT, or ACT scores, though college entrance examinations are not specifically designed for this purpose. The National Merit Scholarships, awarded by the National Merit Scholarship Corporation, are given on the basis of PSAT/NMSQT scores. Roughly 62 percent of National Merit Scholarship are awarded to men. Critics claim that gender, economic, racial, geographic, and language biases in the PSAT/NMSQT place women, the poor, minorities, rural Americans, and individuals whose first or primary language is not English at a competitive disadvantage for these awards.

Standardized testing has become an established feature of American life. More than 100 million standardized tests are administered each year. Despite their flaws, college entrance examinations provide a systematic and relatively objective means to select candidates for admission. For schools that receive large numbers of applications for admission each year, test scores provide a relatively inexpensive and time-effective way to screen candidates. Realizing the flaws in standardized testing and the danger of overreliance on test scores, some institutions have developed alternatives to the traditional selection process.

Bibliography

Buckley, Jack, Letukas, Lynn, and Wildavsky, Ben, editors. Measuring Success: Testing, Grades, and the Future of College Admissions. Johns Hopkins University Press, 2018.

Crouse James and Dale Trusheim, Dale. The Case Against the SAT. Chicago: University of Chicago Press, 1988.

Davey, Charles. The Uses and Misuses of Tests. San Francisco: Jossey-Bass, 1984.

Hoffmann, Banesh. The Tyranny of Testing. New York: Crowell-Collier, 1962.

Jencks, Christopher, and Phillips, Meredith. The Black-White Test Score Gap. Brookings Institution Press, 2011.

Rosser, Phyllis. Sex Bias in College Admissions Tests: Why Women Lose Out. 3d ed., Cambridge, Mass.: The Center, 1989.

“SAT Scores.” National Center for Educational Statistics, 2023, nces.ed.gov/fastfacts/display.asp?id=171. Accessed 13 Jan. 2025

Weiss, John G., Beckwith, Barbara, and Schaeffer, Bob. Standing Up to the SAT. New York: Arco, 1989.